Chang et al. Translational Psychiatry (2020) 10:370 https://doi.org/10.1038/s41398-020-01064-1 Translational Psychiatry

ARTICLE Open Access Mitochondrial DNA and risk of attention deficit and hyperactivity disorder in European Americans Xiao Chang 1, Yichuan Liu1, Frank Mentch1, Joseph Glessner 1,HuiqiQu 1, Kenny Nguyen1, Patrick M. A. Sleiman1,2,3 and Hakon Hakonarson 1,2,3

Abstract Although mitochondrial dysfunction has been implicated in the pathophysiology of attention deficit and hyperactivity disorder ADHD, the role of mitochondrial DNA (mtDNA) has not been extensively investigated. To determine whether mtDNA haplogroups influence risk of ADHD, we performed a case-control study comprising 2076 ADHD cases and 5078 healthy controls, all of whom were European decedents recruited from The Children’s Hospital of Philadelphia (CHOP). Associations between eight major European mtDNA Haplogroups and ADHD risk were assessed in three independent European cohorts. Meta-analysis of the three studies indicated that mtDNA haplogroups K (odds ratio = −4 −3 −4 −3 0.69, P = 2.24 × 10 , Pcorrected = 1.79 × 10 ) and U (odds ratio = 0.77, P = 8.88 × 10 , Pcorrected = 7.11 × 10 ) were significantly associated with reduced risk of ADHD. In contrast, HHV* (odds ratio = 1.18, P = 2.32 × 10−3, Pcorrected = 0.019) was significantly associated with increased risk of ADHD. Our results provide novel insight into the genetic basis of ADHD, implicating mitochondrial mechanisms in the pathophysiology of this relatively common psychiatric disorder. 1234567890():,; 1234567890():,; 1234567890():,; 1234567890():,;

Introduction single-nucleotide variants (SNVs) have also been reported – Attention-deficit hyperactivity disorder (ADHD) is a in multiple ADHD studies3 9. However, most genetic common neurodevelopmental disorder in children and studies of ADHD focused on nuclear DNA (nDNA), and adolescents, characterized by symptoms of hyperactivity, detected susceptibility genes of ADHD so far only account impulsivity, and inattention. It has been recognized as a for a small fraction of the genetic contribution to the highly heritable disorder with a twin-estimated heritability disease. Mitochondrial DNA (mtDNA) therefore emerges around 80%1. Genetic advances in ADHD revealed a as an appealing candidate for the heritability of ADHD. polygenic genetic architecture with multiple common and Mitochondria are best known for their pivotal role in rare genetic factors involved1. Recent genome-wide energy production but are also involved in other cellular association studies (GWAS) have identified common processes, such as amino-acid, lipid and steroid metabo- variants at 12 loci that are significantly associated ADHD lism, and regulation of apoptosis10. Human mitochondria risk through analyzing 20,183 ADHD cases and 35,191 possess a small circular genome, which sequentially controls2. Rare copy number variations (CNVs) and accumulates genetic variants via maternal inheritance. Hence, human mitochondria genomes are classified into several mtDNA haplogroups according to accumulated Correspondence: Xiao Chang ([email protected])or Hakon Hakonarson ([email protected]) mtDNA variants during ancient migrations of human 1Center for Applied Genomics, The Children’s Hospital of Philadelphia, populations11. Previous studies indicate that mtDNA Philadelphia, PA, USA haplogroups display diverse metabolic capacities, and play 2Department of Pediatrics, The Perelman School of Medicine, University of Pennsylvania, Philadelphia, PA, USA an important role in multiple diseases and traits, such as Full list of author information is available at the end of the article

© The Author(s) 2020 Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a linktotheCreativeCommons license, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/. Chang et al. Translational Psychiatry (2020) 10:370 Page 2 of 6

– cancers, neurodegenerative diseases, and longevity12 16. Inclusion criteria for ADHD cases included (1) ages above Recently, cumulative evidence suggested that mitochon- four years old; (2) two or more diagnosis of ICD9 codes drial dysfunction might also lead to a disruption of normal for ADHD (314.1–314.9) in two separate in-person visits neuronal plasticity and promote the development of and more than two prescriptions of ADHD-related psychiatric disorders10,17, implying a potential impact of medications; or (3) three or more ICD9 codes for mtDNA haplogroups in the risk of neuropsychiatric dis- ADHD on separate calendar days. Exclusion criteria for orders, such as ADHD. However, to the best of our ADHD cases were conditions related to brain anomalies knowledge, the association between mtDNA haplogroups or tumors, mental retardation, or psychiatric disorders. and ADHD has been only evaluated in a small Korean Controls were determined based on the following criteria population of 150 cases and 322 controls18. Thus, the (1) ages above four years old; (2) no diagnosis of ICD9 genetic relationship between mtDNA haplogroups and codes for ADHD or other psychiatric, neurological and ADHD has not yet been studied in European populations. related disorders; (3) no mention of the terms “ADHD”, In this study, we investigated the mtDNA haplogroup “attention deficit”, “hyperkinesia”, “hyperkinesis”, “hyper- distribution in three independent European cohorts of kinetic”,or“hyperactivity”; (4) no ADHD related medi- patients with ADHD and compared them with healthy cation. The ADHD case and control selection method was control individuals, aiming to identify potential mtDNA further validated by manual chart review of 100 randomly haplogroups conferring risk to ADHD in European selected cases (N = 50) and controls (N = 50), and yielded populations. positive predictive values (PPVs) >96% for both cases and controls. All three groups were analyzed separately with Methods their array and age/sex, race matched controls, and then Participants meta-analyzed for the study as described in the results This case-control study included three independent section. ADHD cohorts. Blood-derived DNA from participants Written informed consent was obtained from all parti- from the three cohorts were genotyped at the Center for cipants by nursing and medical assistant study staff under Applied Genomics (CAG), at The Children’s Hospital of the direction of CHOP clinicians. All data were analyzed Philadelphia (CHOP), using the Illumina HumanHap550 anonymously and all clinical investigations were conducted (HH550, cohort 1), HumanHap610 (HH610, cohort 2), according to the principles expressed in the Declaration of and Infinium Global Screening (GSA) SNP arrays (GSA, Helsinki. This study was approved by the Research Ethics cohort 3), respectively. Board of CHOP (CHOP IRB#: IRB 06-004886). Cases genotyped of the HH550 chip were recruited from the pediatric and behavioral health clinics at CHOP, Haplogroup assignment capturing the greater Philadelphia area and included mtDNA SNV markers interrogated by the Human- children (age 6–18 y/o) diagnosed with ADHD following Hap550, HumanHap610, and GSA SNP arrays were used the K-SADS (Schedule for Affective Disorders and Schi- in this analysis. There are 163 mtDNA markers included zophrenia for School-Age Children; Epidemiologic Ver- on the HumanHap550 SNP array, 138 mtDNA markers sion) interview. Exclusion criteria were prematurity on the HumanHap610 SNP array, and 131 mtDNA (<36 weeks), intellectual disability, major medical and markers on the GSA SNP array. The coordinates of neurological disorders, pervasive developmental disorder, mtDNA markers in the HumanHap550 and Human- psychoses, and major mood disorders as described pre- Hap610 SNP array were based on the AF347015.1 mito- viously4,19. Controls for the HH550 cohort were recruited chondrial reference genome, which were converted to the by CAG through the CHOP Health Care Network revised Cambridge Reference Sequence (rCRS, GenBank including four primary care clinics and several group NC012920). The mtDNA haplogroup of each participant practices and outpatient practices that performed well- was determined by HaploGrep 220. Given ambiguities in child visits. Healthy children aged 6–18 years old, with no precise haplogroup assignment, evolutionary related serious underlying medical disorder, including but not haplogroups B, F, H, HV, P, R, and V were classified as the limited to neurodevelopmental disorders, cancer, chro- super-haplogroup HHV*; haplogroups O and X were mosomal abnormalities, and known metabolic or genetic classified as the super-haplogroup OX. The accuracy, disorders were included4,19. sensitivity, and specificity of mtDNA haplogroup assign- Individuals genotyped on the HH610 and GSA arrays ment were verified using the three sets of SNVs in 11240 were selected from the biorepository at the CAG, which mitochondrial reference genomes with known hap- collected over 100,000 unique samples recruited from logroups, downloaded from GenBank. Eight haplogroups CHOP from subjects linked to their electronic health common in European populations, including HHV*, I, J, records (EHRs). Cases and controls were selected based K, OX, T, U, and W, were analyzed in this study (Sup- on diagnostic and medication information from EHRs. plementary Table S1). Chang et al. Translational Psychiatry (2020) 10:370 Page 3 of 6

Statistical analysis including 661 cases and 1044 controls (odds ratio 0.7, P = − The association between each haplogroup and ADHD 5.85 × 10 3, Table 1). Meta-analysis combing the asso- risk was tested in R (version 3.5.1) using logistic regres- ciation results from three studies indicated that both sion analyses. Gender and the first ten principal compo- haplogroup K and U yielded a P value that extends well nents were included as covariates. Principal component beyond the significance threshold after adjustment for analysis (PCA) was performed using EIGENSTRAT to multiple testing (Table 2). In addition, haplogroup HHV*, infer genetic ancestry of samples and detect population the most frequent haplogroup in European populations, substructures21. Participants with European ancestry were was significantly associated with increased risk of ADHD − selected by comparing with four reference European after meta-analysis (odds ratio 1.176, P = 2.32 × 10 3, populations from 1000 Genome, including Utah Residents Pcorrected = 0.0185, Table 2). with Northern and Western European ancestry (CEU), Toscani in Italia (TSI), British in England and Scotland Discussion (GBR), and Iberian Population in (IBS). PCA plots To our knowledge, this is the first large-scale study to of the population structures are provided in Supplemen- investigate the role of mtDNA haplogroups in ADHD in tary Fig. S1. Meta-analysis of summary statistics from the European populations. Our findings indicate that hap- three cohorts was performed by PLINK (version 1.9)22. logroups U and K are associated with reduced risk of Fixed effects P values were reported. Bonferroni correc- ADHD in patients of European descent, while the super- tion was used for correction of multiple testing. A cor- haplogroup HHV* is associated with increased risk of rected P < 0.05 was considered statistically significant. The ADHD. Our findings demonstrate that genetic variations kinship between pairs of subjects was detected by identity of mtDNA are implicated in the development of ADHD. by descent (IBD) analysis, and one of the related indivi- Although associations between mtDNA haplogroups and duals with proportional IBD > 0.25 were excluded. ADHD have been reported in a recent study of Korean children18,thesefindings need to be further verified by Results studies of larger sample sizes and other ethnicity popula- The HH550 cohort comprising 764 European-American tions. In this study, the association between mtDNA hap- ADHD cases and 2031 ancestry-matched controls was logroups and ADHD was consistently observed in three used in the discovery analysis. Haplogroups including I, J, independent European populations, adding robust evidence K, L, T, W, OX, and HHV* were tested, as they were to support the role of mtDNA haplogroups in ADHD. mainly presented in European populations. In consistent Cases and controls of the HH550 cohort used here were with this notion, more than 97% of individuals in the previously investigated in CNV and GWAS studies of HH550 cohort belonged to the eight common European ADHD4,19, which successfully identified multiple genetic haplogroups. In addition, the HHV* haplogroup encom- risk factors of ADHD. Cases and controls from HH610 and passed about 50% of subjects in the HH550 cohort, which GSA cohorts were determined based on recurrent use of was consistent with previous mtDNA studies of European diagnostic codes and medication information from EHRs. populations23,24. Association analyses indicated that hap- The accuracy of this selection method was further validated logroup K was significantly less represented in patients with high accuracy by manual chart review. Overall, cases with ADHD than healthy controls (odds ratio 0.64, P = and controls were appropriately selected for this case- 0.00443, Pcorrected = 0.0354, Table 1), suggesting a pro- control study. European subjects were strictly selected tective role of haplogroup K in ADHD risk. based on their genetic ancestry inferred from genotyping Associations between eight mtDNA haplogroups and array data. Association results were further adjusted by the ADHD risk were also assessed in an independent HH610 first ten PCs calculated from the genotyping array data in cohort containing 651 cases and 2003 controls. A sig- order to reduce the possibility of spurious association nificant association between Haplogroup K and ADHD caused by population stratification, risk was observed (odds ratio 0.66, P = 0.0105, Table 1). A protective role of haplogroups K and U was observed Interestingly, haplogroup U, from which haplogroup K in this study. Although they are classified as two major was derived, was also significantly associated with reduced haplogroups in European populations, haplogroup K was risk of ADHD (odds ratio 0.71, P = 0.0233, Table 1). We recognized as a subclade of haplogroup U, implying those therefore combined haplogroups U and K as a super- two haplogroups may have similar mitochondrial activities haplogroup UK, and detected a significant association and functions. Consistent with this, haplogroups K and U after adjustment for multiple testing between the UK displayed an uncoupling effect leading to reduced adeno- haplogroup and ADHD risk (odds ratio 0.66, P = 3.08 × sine triphosphate (ATP) production and energy genera- −4 −3 15,16 10 , Pcorrected = 2.46 × 10 , Table 1). tion . Cybrids of haplogroups K and U also displayed The association between the super-haplogroup UK and lower inner mitochondrial membrane potential, mito- ADHD risk was further replicated in another cohort GSA chondrial protein synthesis, and cell growth capacity than Chang et al. Translational Psychiatry (2020) 10:370 Page 4 of 6

Table 1 Associations of eight haplogroups in three European cohorts.

Study Haplogroup Ncase Ncontrol Fcase Fcontrol OR (95% CI) PPcorrected

HH550 OX 8 34 1.05% 1.67% 0.62 (0.29,1.35) 0.244 1 W 17 31 2.23% 1.53% 1.47 (0.81,2.67) 0.268 1 K 50 199 6.54% 9.80% 0.64 (0.47,0.89) 4.43 × 10−3 0.035 U 94 292 12.30% 14.38% 0.84 (0.65,1.07) 0.197 1 J 71 204 9.29% 10.04% 0.92 (0.69,1.22) 0.653 1 T 74 184 9.69% 9.06% 1.08 (0.81,1.43) 0.653 1 I 30 58 3.93% 2.86% 1.39 (0.89,2.18) 0.13 1 HHV* 396 972 51.83% 47.86% 1.17 (0.99,1.38) 0.071 0.564 UK 144 491 18.85% 24.18% 0.73 (0.59,0.9) 2.63 × 10−3 0.021 HH610 OX 7 32 1.08% 1.60% 0.67 (0.29,1.52) 0.349 1 W 10 32 1.54% 1.60% 0.96 (0.47,1.97) 0.865 1 K 43 193 6.61% 9.64% 0.66 (0.47,0.93) 0.011 0.084 U 73 304 11.21% 15.18% 0.71 (0.54,0.93) 0.023 0.186 J 73 197 11.21% 9.84% 1.16 (0.87,1.54) 0.269 1 T 70 196 10.75% 9.79% 1.11 (0.83,1.48) 0.437 1 I 17 61 2.61% 3.05% 0.85 (0.5,1.47) 0.684 1 HHV* 337 940 51.77% 46.93% 1.21 (1.02,1.45) 0.086 0.69 UK 116 497 17.82% 24.81% 0.66 (0.52,0.82) 3.08 × 10−4 2.46 × 10−3 GSA OX 0 2 0.00% 0.19% 0 (0,Inf) 0.972 1 W 0 0 0.00% 0.00% NA NA NA K 46 98 6.96% 9.39% 0.72 (0.5,1.04) 0.098 0.785 U 78 162 11.80% 15.52% 0.73 (0.55,0.97) 0.046 0.368 J 55 88 8.32% 8.43% 0.99 (0.69,1.4) 0.64 1 T 66 98 9.98% 9.39% 1.07 (0.77,1.49) 0.574 1 I 19 32 2.87% 3.07% 0.94 (0.53,1.67) 0.739 1 HHV* 350 500 52.95% 47.89% 1.22 (1.01,1.49) 0.035 0.278 UK 124 260 18.76% 24.90% 0.7 (0.55,0.89) 5.85 × 10−3 0.047 haplogroup H25. Hence, the protective effect of haplogroup higher ATP levels and oxidative phosphorylation K and U could be partially explained by energy deficiency OXPHOS capacities25,32. It has also been reported to and/or reduced load of harmful reactions. In addition, increase the risk of Alzheimer’s disease and lower the age members of haplogroups K and U represented higher brain onset of Huntington disease presentation29,32. pH than others26. Higher pH is correlated with a lower This study has two limitations. First, the mtDNA hap- brain activity and confers protection against psychiatric logroups were deduced from genotyping arrays. It disorders27. In fact, haplogroup K and U have also been remains unclear which precise genetic variants are causal implicated in multiple neurological and neurodegenerative in ADHD. This limitation can be solved in future studies disorders in Europeans, such as transient ischemic attack, by sequencing the whole mtDNA genome in a large col- ischemic stroke, neuroblastoma, Parkinson’s disease, and lection of ADHD cases and controls. Second, the three – Alzheimer’s disease23,24,28 31. study cohorts were genotyped at the same location, and it Our study also revealed that the HHV* haplogroup, the is possible that unknown potential confounding factors most dominant haplogroup in European populations, is might have influenced the results. Third, the results could associated with increased risk of ADHD. In contrast to be confounded by maternal ancestry, since human haplogroup K and U, haplogroup H was associated with mtDNA is maternally inherited. However, this effect is Chang et al. Translational Psychiatry (2020) 10:370 Page 5 of 6

Table 2 Meta-analysis of results from three European References cohorts. 1. Faraone, S. V. & Larsson, H. Genetics of attention deficit hyperactivity disorder. Mol. Psychiatry 24,562–575 (2019). fi fi HaploGroup OR PPI 2. Demontis, D. et al. Discovery of the rst genome-wide signi cant risk loci for corrected attention deficit/hyperactivity disorder. Nat. Genet. 51,63–75 (2019). 3. de Silva, M. G. et al. Disruption of a novel member of a sodium/hydrogen OX 0.63 0.119 0.953 0 exchanger family and DOCK3 is associated with an attention deficit hyper- – W 1.18 0.478 1 0 activity disorder-like phenotype. J. Med. Genet. 40,733 740 (2003). 4. Elia, J. et al. Genome-wide copy number variation study associates metabo- − − K 0.69 2.24 × 10 4 1.79 × 10 3 0 tropic glutamate receptor gene networks with attention deficit hyperactivity – −4 −3 disorder. Nat. Genet. 44,78 84 (2011). U 0.77 8.88 × 10 7.11 × 10 0 5. Jarick, I. et al. Genome-wide analysis of rare copy number variations reveals fi J 1.05 0.611 1 0 PARK2 as a candidate gene for attention-de cit/hyperactivity disorder. Mol. Psychiatry 19,115–121 (2014). T 1.08 0.404 1 0 6. Kim, D. S. et al. Sequencing of sporadic Attention-Deficit Hyperactivity Dis- order (ADHD) identifies novel and potentially pathogenic de novo variants I 1.09 0.586 1 3.29 and excludes overlap with genes associated with autism spectrum disorder. − – HHV* 1.18 2.32 × 10 3 0.019 0 Am. J. Med. Genet. B Neuropsychiatr. Genet. 174,381 389 (2017). 7. Lionel,A.C.etal.Rarecopynumbervariation discovery and cross-disorder − − UK 0.71 1.94 × 10 7 1.55 × 10 6 0 comparisons identify risk genes for ADHD. Sci. Transl. Med. 3, 95ra75 (2011). 8.Connolly,J.J.,Glessner,J.T.,Elia,J.&Hakonarson,H.ADHD&pharma- I: I2 heterogeneity index (0–100) cotherapy: past, present and future: a review of the changing landscape of drug therapy for attention deficit hyperactivity disorder. Ther.Innov.Regul.Sci. 49,632–642 (2015). not likely to be significant as it is much smaller for the 9. Elia, J. et al. Fasoracetam in adolescents with ADHD and glutamatergic gene homogeneous European population than for admixed network variants disrupting mGluR neurotransmitter signaling. Nat. Commun. 9, 4 (2018). populations. 10. Manji, H. et al. Impaired mitochondrial function in psychiatric disorders. Nat. Rev. Neurosci. 13, 293–307 (2012). Conclusion 11. Wallace,D.C.MitochondrialDNAvariation in human radiation and disease. Cell 163,33–38 (2015). This study demonstrates that mtDNA haplogroups K 12. Bilal, E. et al. Mitochondrial DNA haplogroup D4a is a marker for extreme and U confer protective effects against ADHD, while longevity in Japan. PLoS ONE 3, e2421 (2008). mtDNA haplogroup HHV* is associated with increased 13. Castri, L. et al. A mitochondrial haplogroup is associated with decreased longevity in a historic new world population. Hum. Biol. 86,251–259 (2014). risk of ADHD, providing new insights into the genetic 14. Cocos, R., Schipor, S., Badiu, C. & Raicu, F. Mitochondrial DNA haplogroup K as basis of ADHD. A better understanding of the role of a contributor to protection against thyroid cancer in a population from mitochondria in the pathogenesis of ADHD might lead to southeast . Mitochondrion 39,43–50 (2018). 15.Niemi,A.K.&Majamaa,K.MitochondrialDNAandACTN3genotypesin new methods for the prevention and treatment of this Finnish elite endurance and sprint athletes. Eur. J. Hum. Genet. 13,965–969 common disease. (2005). 16. Tranah, G. J. et al. Mitochondrial DNA variation in human metabolic rate and Acknowledgements energy expenditure. Mitochondrion 11,855–861 (2011). We thank the patients and their families for their participation in this study. 17. Verma, P. et al. Attention deficit-hyperactivity disorder suffers from mito- This work was supported by Institutional Development Funds from the CHOP chondrial dysfunction. BBA Clin. 6, 153–158 (2016). (H.H.) and by NIH grants U01HG006830 (H.H.). 18. Hwang, I. W. et al. Assessment of associations between mitochondrial DNA haplogroups and attention deficit and hyperactivity disorder in Korean chil- dren. Mitochondrion 47, 174–178 (2019). Author details 19. Elia, J. et al. Rare structural variants found in attention-deficit hyperactivity 1 Center for Applied Genomics, The Children’s Hospital of Philadelphia, disorder are preferentially associated with neurodevelopmental genes. Mol. 2 Philadelphia, PA, USA. Department of Pediatrics, The Perelman School of Psychiatry 15,637–646 (2010). 3 Medicine, University of Pennsylvania, Philadelphia, PA, USA. Division of Human 20. Weissensteiner, H. et al. HaploGrep 2: mitochondrial haplogroup classification Genetics, Department of Pediatrics, The Perelman School of Medicine, in the era of high-throughput sequencing. Nucleic Acids Res. 44,W58–W63 University of Pennsylvania, Philadelphia, PA, USA (2016). 21. Price, A. L. et al. Principal components analysis corrects for stratification in Conflict of interest genome-wide association studies. Nat. Genet. 38,904–909 (2006). The authors declare that they have no conflict of interest. 22. Purcell, S. et al. PLINK: a tool set for whole-genome association and population-based linkage analyses. Am.J.Hum.Genet.81, 559–575 (2007). 23. Lakatos, A. et al. Association between mitochondrial DNA variations and Publisher’s note Alzheimer’s disease in the ADNI cohort. Neurobiol. Aging 31,1355–1363 (2010). Springer Nature remains neutral with regard to jurisdictional claims in 24. Chinnery, P. F., Elliott, H. R., Syed, A., Rothwell, P. M. & Oxford Vascular, S. published maps and institutional affiliations. Mitochondrial DNA haplogroups and risk of transient ischaemic attack and ischaemic stroke: a genetic association study. Lancet Neurol. 9,498–503 (2010). Supplementary Information accompanies this paper at (https://doi.org/ 25. Gomez-Duran, A. et al. Unmasking the causes of multifactorial disorders: 10.1038/s41398-020-01064-1). OXPHOS differences between mitochondrial haplogroups. Hum. Mol. Genet. 19,3343–3353 (2010). 26. Rollins, B. et al. Mitochondrial variants in schizophrenia, bipolar disorder, and Received: 1 March 2020 Revised: 13 September 2020 Accepted: 5 October major depressive disorder. PLoS ONE 4, e4913 (2009). 2020 27. Hagihara, H. et al. Decreased brain pH as a shared endophenotype of psy- chiatric disorders. Neuropsychopharmacology 43,459–468 (2018). Chang et al. Translational Psychiatry (2020) 10:370 Page 6 of 6

28. Ghezzi,D.etal.MitochondrialDNAhaplogroup K is associated with a lower 31. Chang, X. et al. Mitochondrial DNA haplogroups and susceptibility to neuro- risk of Parkinson’s disease in Italians. Eur. J. Hum. Genet. 13,748–752 (2005). blastoma. J. Natl Cancer Inst. djaa024, https://doi.org/10.1093/jnci/djaa024. 29. Maruszak, A. et al. The impact of mitochondrial and nuclear DNA variants on (2020). Online ahead of print. late-onset Alzheimer’s disease risk. J. Alzheimers Dis. 27,197–210 (2011). 32. Arning, L. et al. Mitochondrial haplogroup H correlates with ATP levels and 30. Pyle, A. et al. Mitochondrial DNA haplogroup cluster UKJT reduces the risk of age at onset in Huntington disease. J. Mol. Med. 88,431–436 (2010). PD. Ann. Neurol. 57,564–567 (2005).